PROJECT SUMMARY/ABSTRACT Reading is an important development in human brain function, and it is an integral part of how we interact with other people. The impairment or loss of reading (alexia) due to a stroke can be devastating, yet an adequate understanding of how the brain processes reading functions is still lacking. From what is currently known, cognitive models of reading describe how pathways between orthographic, phonologic and semantic processing contribute to successful reading. And features of words, such as concreteness, regularity, and frequency, play an important role in identifying these pathways. Reading behaviors have also been related to specific brain areas. However, there is little literature that relates the proposed cognitive routes to neural networks. This proposal aims to connect these two fields by examining how the cognitive components of reading relate to the neurobiological processes that perform them (Aim 1). Furthermore, when the reading system breaks down, the precise mechanisms that contribute to residual reading abilities are not well specified. Deficits are usually defined by the words that cannot be read, or by the location of the damage done by a stroke or other trauma. This proposal aims to determine the role that the remaining function of the neural reading network plays in defining deficient reading behavior (Aim 2). In the proposed study, participants will undergo reading-task related fMRI scans, analyzed by functional connectivity (FC) and psychophysiological interaction (PPI) analyses, so that their reading-related connectivity can be assessed. Functional localizer tasks will be used to identify areas in the brain that govern cognitive reading components. A general reading task, with feature-controlled words, will be used for a reading-related FC analysis between those brain areas to identify reading pathways. To identify reading-related neural pathways, FC analyses will first evaluate the correlation between the activity of the identified areas; then PPI analyses will evaluate the correlation between FC changes and reading words with features that utilize specific cognitive pathways. These methods will be used to test the hypotheses that: a typically reading, older population will show reading-related FC between the areas canonically associated with reading processing (Aim 1a), and reading words that have high vs. low values of specific word features will produce differing effects in connections that reflect phonologic and semantic pathways (Aim 1b). The FC maps of pairs of participants whose lesions have significant overlap but who have different reading profiles will be compared. This will test the hypothesis that functional networks will explain variance in behavioral deficits (Aim 2). The results of this study should provide new insights into the fundamental questions of how cognitive reading routes relate to the reading network pathways in the brain, and how the surviving neural network contributes to residual reading post-stroke, which can be applied to the development of novel treatments for alexia.